1. Field of the Invention
The present invention relates to image processing apparatuses and in particular to image processing apparatuses capable of appropriate color correction with reduced memory capacity.
2. Description of the Related Art
In the field of color printing and color copying, there have been known image processing apparatuses by which image data read in the RGB system are output as data in the CYMK system. In such an image processing apparatus, to reduce its memory capacity a lookup table having a reduced capacity and linear interpolation are used to divide each input signal by a predetermined interval and apply linear interpolation in the divided regions.
In conventional color corrections, colors are previously corrected considering the reproduction characteristics of the colors of the toner in the printing system to reproduce the colors faithfully. Such techniques include those based on linear and non-linear color-masking techniques, interpolation technique using a lookup table (referred to as a LUT hereinafter), and the like. According to interpolation technique using the LUT, the relation between an input value and an output value is free of any limitations and any non-linear curves can be represented so that errors due to color correction can be reduced. The size of the LUT can also be changed depending on the object and demand.
An interpolation technique using a LUT will now be described with reference to the block diagram shown in FIG. 9. There will be described here a system in which an input color space is assumed to be signals C0, M0, Y0 in a CMY color space converted from an RGB color space via various processes, such as tone conversion, UCR/BP, and its output color space after an interpolation process is assumed to be data C, M, Y considering the color reproductivity of the toner of the printing system. The system may of course be used for another single type of color-space conversion, such as that from an RGB color space to a YCrCb color space, that from an RGB color space to a CIEL*a*b* color space.
Referring to FIG. 9, a conventional color-correction block includes a storage table 301 receiving signals C0, M0, Y0 of the CMY color space, and an interpolator 302 using data output and input from and to storage table 301 to interpolate each data.
Storage table 301 is a three-dimensional LUT for storing a plurality of data C, M, Y to be output for input signals C0, M0, Y0, typically dividing each of the three orthogonal axes of the CMY color space into N regions at predetermined intervals to store (N+1)3 even representative terms, wherein N represents an integer. In general N=8 is often applied, 729 even representative terms designated to form the LUT. The three-dimensional LUT stores a combination of  less than C0i, M0i, Y0i greater than  and  less than Ci, Mi, Yi greater than  of each lattice point, wherein i represents an integer from 0 to Nxe2x88x921. It is determined to which equally divided region any signals C0, M0, Y0 input each belong, and the  less than C0i, M0i, Y0i greater than  and  less than Ci, Mi, Yi greater than  of a lattice point constituting the region are output to the subsequent stage or interpolator 302.
In interpolator 302, a unit cube with each lattice point as a vertex is constructed from lattice point data  less than C0i, M0i, Y0i greater than  and  less than Ci, Mi, Yi greater than  selected by input signals C0, M0, Y0 and output from storage table 301. Input signals C0, M0, Y0 are interpolated according to any of the interpolation methods described below, depending on the positional relation in the unit cube, to produce and output output signals C, M, Y:
(a) CUBE method: an interpolation method using eight vertices of a unit cube, the interpolation expression being an expression of eight terms of degree three.
(b) TETRA HEDRON method: an interpolation method which divides a unit cube into six tetrahedra having an equal volume around a diagonal axis to use four vertices thereof, the interpolation expression being an expression of four terms of degree one.
(c) PYRAMID method: an interpolation method which divides a unit cube into three pyramids having an equal volume around a diagonal axis of the unit cube to use five vertices thereof, the interpolation expression being an expression of five terms of degree three.
(d) PRISM method: an interpolation method which bisects a unit cube along a plane parallel to an axis (e.g. the C axis) to divide it into two triangular prisms to use six vertices thereof, the interpolation expression being an expression of six terms of degree two.
As described above, interpolation methods using a three-dimensional LUT are essentially based on the concept that a non-linear curve in a color space as a complex model can be divided into sufficiently small regions to establish a linear relation in the segment and allow approximation. For conventional color correction techniques, however, lattice point data  less than C0i, M0i, Y0i greater than  and  less than Ci, Mi, Yi greater than  stored in the three-dimensional LUT are configured of data each axis for which is divided equally in N, ignoring an original, non-linear curve provided by a relation between all input and output signals. Consequently, depending on the shape of a non-linear curve, e.g. a jagged curve with a plurality of inflection points, a linear relation is not established in a divided region and an approximation error between non-linearity and linearity will be increased disadvantageously.
FIG. 10 illustrates an example of such disadvantage. In this figure, linear interpolation of a three-dimensional color space is represented in one dimension (an expression of one variable) of C=f(x) for simplicity. Also for simplicity, N is equal to four and an input signal is equally divided into 0, 63, 127, 192, 255, and these values are represented along the x axis and the f(x) at each point is plotted along the x axis.
In this example, the non-linear curve obtained from all input values and output values obtained therefrom is shaped to have two inflection points, as shown in a narrow line a whereas the straight lines binding the output values at points 0, 63, 127, 192, 255 obtained by the equal division of the input signal are represented by a wide line b. Thus, particularly in such divided regions as those between 0 and 63, and between 127 and 192, the error between the non-linear curve (narrow line a) and the linearly approximate, straight line (wide line b) increases and a faithful result of interpolation cannot be obtained.
One object of the present invention is therefore to obtain appropriate output data corresponding to input data with less data than that of the input data when the input and output data are of non-linear.
Another object of the present invention is to provide an image processing method capable of using reduced memory capacity to precisely modify an output signal.
Still another object of the present invention is to provide an image processing apparatus capable of using reduced memory capacity to precisely modify an output signal.
The above objects of the present invention are achieved in accordance with the image processing method including the following elements. More specifically, in accordance with the present invention a data processing method outputting color data in non-linear relation with input data includes the steps of: forming M color images depending on M input data; reading the M formed color images to obtain M color data; picking up N sets of data including a set of input data and color data close to an inflection point, depending on a relation between the M input data and the M color data, wherein M greater than N; and performing operation based on the N sets of data to obtain output data corresponding to the input data.
Since N sets of data including a set of input data and color data close to an inflection point can be picked up from a relation between M input data and M color data, wherein M greater than N, and the N data can be used to obtain output data corresponding to the input data, the output data can be appropriately obtained corresponding to the input data using data less than the input data when the input data and the output data are non-linear.
In another aspect of the present invention, a data processing apparatus outputting color data in non-linear relation with input data includes: a color-image formation portion for forming M color images depending on M input data; a reception portion reading the M formed color images to obtain M color data; a picking-up portion for picking up N sets of data including a set of input data and color data close to an inflection point, depending on a relation between the M input data and the M color data, within M greater than N; and an operation portion for performing anoperation based on the N sets of data to obtain output data corresponding to the input data.
Since N sets of data including a set of input data and color data close to an inflection point can be picked up from a relation between M input data and M color data, wherein M greater than N, and the N data can be used to obtain output data corresponding to the input data, there can be provided a data processing apparatus capable of using reduced storage capacity to precisely modify an output signal.
Preferably, the picking-up portion includes: difference-value calculation portion for calculating a difference between each of M color data received and by the reception portion and each of two color data adjacent thereto and a select portion for selecting N difference values of a plurality of such obtained difference values in descending order.
The difference value between any adjacent two of M color data is calculated and N data are determined among such difference values in descending order. Thus data to be stored can readily be determined.
Preferably, the pick-up portion has previously stored input data and a targeted corrected value to pick up N color depending on a result obtained from input data and on the targeted corrected value.
The foregoing and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.